Delays and Timetabling for Passenger Trains CARL-WILLIAM PALMQVIST FACULTY OF ENGINEERING | LUND UNIVERSITY 2019 317 Travel by train has increased steadily for the last 30 years. In order to build trust in and shift even more traffic to railways, more trains must arrive on time. In practice, many train delays are caused by small disturbances at stations, which add up. One issue is that the scheduled dwell times are simply too short. Another is that punctuality falls quickly if it is either warm or cold. A third is that interactions between trains rarely go as planned. One suggestion for how to reduce delays is to paint markings that show where passengers should wait. Another is to remove switches, so that remaining ones can be better maintained. A third is to make railways more resilient to the weather variations of today, and to the climate changes of tomorrow. Cost-effective improvements can also be made with timetables. More of the planning can be automated, so that planners can focus more on setting appropriate dwell times and on improving the timetabling guidelines. This way, many more trains can be on time. Lund University Faculty of Engineering Department of Technology and Society Bulletin 317 953103 ISBN 978-91-7895-310-3 ISSN 1653-1930 789178 CODEN: LUTVDG/(TVTT-1059)1-105 9 317 Delays and Timetabling for Passenger Trains Delays and Timetabling for Passenger Trains Carl-William Palmqvist DOCTORAL DISSERTATION by due permission of the Faculty of Engineering, Lund University, Sweden. To be defended at V:C, V-House. John Ericssons väg 1, Lund. 2019-11-08 at 10:15. Faculty opponent Birre Nyström Organisation Document name DOCTORAL DISSERTATION LUND UNIVERSITY Date of issue 2019-11-08 Carl-William Palmqvist Sponsoring organisation K2 – The Swedish Knowledge Centre for Public Transport Title and subtitle Delays and Timetabling for Passenger Trains Abstract Travel by train has increased steadily in Sweden the last 30 years. The pace has been about two to three percent per year, and we now have twice as many passengers. With growing awareness of the changing climate, the pace is increasing further. A problem that affects both passengers and businesses in Sweden is train delays. One way to describe these is as the share of trains that arrive less than six minutes delayed. About 90% of trains in Sweden meet this standard and have done so for many years. In a way this is impressive, since there are now many more trains. Unfortunately, this also means that more and more passengers are affected by delays. This leads to irritation, threatens the shift of traffic to railways, and costs a lot for society. More trains must arrive on time. This thesis shows that delays are mostly caused by small disturbances – up to a minute or two. Over long journeys, these small disturbances accumulate and sometimes cause quite big delays. These delays mostly occur at stations, where the trains stop, but are then unable to continue on time. It is difficult to say exactly what causes these small disturbances, but the time that the trains are supposed to be at stations – the dwell times – are often too short. Another pattern is seen between delays and weather: if it is either warm or cold, delays increase rapidly. And while winter and snow return every year, they still cause major disruptions. The thesis holds a few suggestions to reduce delays. One is platform markings that show where the trains will stop, where the doors will be, and where the passengers should wait. This is an easy and affordable way to speed up the stops, so that the trains depart on time. Another measure is to remove switches. Then there are fewer parts that can fail, and those that remain can be maintained to a higher standard. A third way is to adapt the railway, so that it better withstands the weather variations of today, and the climate changes of tomorrow. Something that has been done in other countries is to shade and air-condition electronics and signals along the railway. Then the components to not overheat, and more trains run on time. Many things can also be done with timetables, so that more trains run on time, without a rise in costs. More of the planning can be automated. Then more time can be spent on giving trains appropriate dwell times. Infrastructure managers should also do more to evaluate and improve the rules and guidelines that govern timetabling. In this way we can improve timetables gradually from year to year, with fewer and fewer delays as a consequence. These suggestions do not solve all of the railway’s issues, but they would lead to many more trains arriving on time. Key words Railways, Trains, Delays, Punctuality, Timetable Classification system and/or index terms (if any) Supplementary bibliographical information Language English ISSN and key title 1653-1930 ISBN 978-91-7895-310-3 (printed) Bulletin – Lund University, Faculty of Engineering, Department of ISBN 978-91-7895-311-0 (pdf) Technology and Society, 317 Recipient’s notes Number of pages 105 Price Security classification I, the undersigned, being the copyright owner of the abstract of the above-mentioned dissertation, hereby grant to all reference sources permission to publish and disseminate the abstract of the above-mentioned dissertation. Signature Date 2019-09-30 Delays and Timetabling for Passenger Trains Carl-William Palmqvist Front cover photo by Carl-William Palmqvist Back cover photo by Kennet Ruona Copyright pp 1-105 (Carl-William Palmqvist) Paper 1 © WIT Press. This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 United States License. Paper 2 © by the Authors. This is an open-access article distributed under the terms of the Creative Commons Attribution 3.0 United States License. Paper 3 © by the Authors. Paper 4 © by the Authors. This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 United States License. Paper 5 © by the Authors. Faculty of Engineering Technology and Society ISBN 978-91-7895-310-3 (printed) ISBN 978-91-7895-311-0 (pdf) ISSN 1653-1930 Printed in Sweden by Media-Tryck, Lund University Lund 2019 To Irene Table of Contents Populärvetenskaplig sammanfattning ................................................................ 11 Acknowledgements .............................................................................................. 12 Introduction .......................................................................................................... 17 Structure of thesis ......................................................................................... 18 Background ........................................................................................................... 19 Timetable planning ....................................................................................... 19 Quality of timetables ........................................................................... 19 Capacity allocation .............................................................................. 24 Robust timetable planning ............................................................................ 26 Run time supplements ......................................................................... 27 Dwell times .......................................................................................... 28 Headway times .................................................................................... 29 Train delays and punctuality ........................................................................ 31 Delay definitions ................................................................................. 31 Earlier research on delays .................................................................... 33 Research gaps ............................................................................................... 37 Aim ......................................................................................................................... 39 Research questions .............................................................................. 39 Delimitations ....................................................................................... 40 Method and Data .................................................................................................. 41 Quantitative, qualitative and mixed methods ............................................... 41 Working with delays and punctuality ........................................................... 42 Measuring and observing delays ......................................................... 42 Varying punctuality definitions ........................................................... 43 Quantitative data used .................................................................................. 46 Train movement data ........................................................................... 46 Dwell times and passenger counts ....................................................... 48 Timetable data ..................................................................................... 50 Weather data ........................................................................................ 51 Infrastructure data ................................................................................ 52 Analytical methods used .............................................................................. 54 Pre-processing the data ........................................................................ 56 Visual analysis and Welch’s t-test ......................................................